Counting people and making people count: Key sources of population projections

PDF - 477 KB
Helpdesk Query:

What are the key sources of information on population dynamics globally and for the countries and regions where DFID works?


This literature review is the first section of a two part literature review in to population dynamics focussing on a review of key global information sources. Three sources of centrally generated projections exist: The Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, the World Bank and the Wittgenstein Centre. As the World Bank relies heavily on the UN its role will not be considered. The premier source of population projections is the UN Population Division. This data is easily accessible and updated every two years. Five global datasets are produced by the Division (i) World Population Prospects, (ii) World Urbanisation Prospects, (iii) World Population Policies, (iv) Estimates of family planning indicators and (v) Estimates of the number of international migrants.

The 2015 Revision of World Population Prospects is the twenty-fourth round of official UN’s population estimates and projections. Key demographic indicators can be accessed for selected periods or dates from 1950 to 2100, for the world, development groups, major areas, regions, and countries or areas with more than 90,000 inhabitants in 2015. For countries or areas with fewer than 90,000 inhabitants in 2015, only figures related to population size and growth are provided. Data tables and figures for 1950-2015 are estimates and those thereafter are projections. Population projections are essential in the assessment of future need. A projection is
a prediction made by extrapolating from past observations and shows what may happen if a set of assumptions holds true. The assumptions can be varied to see how much difference they make, thus producing a set of variant projections. These could include ‘best-case’ and ‘worstcase’ scenarios.

The projections are presented for medium, high, low and constant-fertility variants. The main results are presented in a series of Excel data files, maps, graphs and publications. Data can be displayed for both sexes combined, male or female. At a regional level and over a period of a few decades, the UN projections have a good track record of predictive validity (1). Figures for specific countries though will be subject to greater error. The mean absolute error of the UN 1994 projections for the size of national populations in the year 2000 was 5%. The predictive accuracy of national projections depends on: (1) accuracy of baseline data (typically a census), ability to detect and adjust for errors; (2) correct mortality assumptions, usually not an important source of error; (3) correct fertility assumptions, a more important source of error; and (4) migration assumptions, which are intrinsically more difficult to forecast than mortality and fertility, but are less problematic for large populations than small ones. Broadly, it is fair to claim that UN projections are of sufficient quality for many planning purposes over a horizon of 10 to 20 years.

PDF - 477 KB

You may also be interested in